xref: /openbmc/linux/fs/btrfs/async-thread.c (revision 9726bfcd)
1 // SPDX-License-Identifier: GPL-2.0
2 /*
3  * Copyright (C) 2007 Oracle.  All rights reserved.
4  * Copyright (C) 2014 Fujitsu.  All rights reserved.
5  */
6 
7 #include <linux/kthread.h>
8 #include <linux/slab.h>
9 #include <linux/list.h>
10 #include <linux/spinlock.h>
11 #include <linux/freezer.h>
12 #include "async-thread.h"
13 #include "ctree.h"
14 
15 #define WORK_DONE_BIT 0
16 #define WORK_ORDER_DONE_BIT 1
17 #define WORK_HIGH_PRIO_BIT 2
18 
19 #define NO_THRESHOLD (-1)
20 #define DFT_THRESHOLD (32)
21 
22 struct __btrfs_workqueue {
23 	struct workqueue_struct *normal_wq;
24 
25 	/* File system this workqueue services */
26 	struct btrfs_fs_info *fs_info;
27 
28 	/* List head pointing to ordered work list */
29 	struct list_head ordered_list;
30 
31 	/* Spinlock for ordered_list */
32 	spinlock_t list_lock;
33 
34 	/* Thresholding related variants */
35 	atomic_t pending;
36 
37 	/* Up limit of concurrency workers */
38 	int limit_active;
39 
40 	/* Current number of concurrency workers */
41 	int current_active;
42 
43 	/* Threshold to change current_active */
44 	int thresh;
45 	unsigned int count;
46 	spinlock_t thres_lock;
47 };
48 
49 struct btrfs_workqueue {
50 	struct __btrfs_workqueue *normal;
51 	struct __btrfs_workqueue *high;
52 };
53 
54 static void normal_work_helper(struct btrfs_work *work);
55 
56 #define BTRFS_WORK_HELPER(name)					\
57 noinline_for_stack void btrfs_##name(struct work_struct *arg)		\
58 {									\
59 	struct btrfs_work *work = container_of(arg, struct btrfs_work,	\
60 					       normal_work);		\
61 	normal_work_helper(work);					\
62 }
63 
64 struct btrfs_fs_info *
65 btrfs_workqueue_owner(const struct __btrfs_workqueue *wq)
66 {
67 	return wq->fs_info;
68 }
69 
70 struct btrfs_fs_info *
71 btrfs_work_owner(const struct btrfs_work *work)
72 {
73 	return work->wq->fs_info;
74 }
75 
76 bool btrfs_workqueue_normal_congested(const struct btrfs_workqueue *wq)
77 {
78 	/*
79 	 * We could compare wq->normal->pending with num_online_cpus()
80 	 * to support "thresh == NO_THRESHOLD" case, but it requires
81 	 * moving up atomic_inc/dec in thresh_queue/exec_hook. Let's
82 	 * postpone it until someone needs the support of that case.
83 	 */
84 	if (wq->normal->thresh == NO_THRESHOLD)
85 		return false;
86 
87 	return atomic_read(&wq->normal->pending) > wq->normal->thresh * 2;
88 }
89 
90 BTRFS_WORK_HELPER(worker_helper);
91 BTRFS_WORK_HELPER(delalloc_helper);
92 BTRFS_WORK_HELPER(flush_delalloc_helper);
93 BTRFS_WORK_HELPER(cache_helper);
94 BTRFS_WORK_HELPER(submit_helper);
95 BTRFS_WORK_HELPER(fixup_helper);
96 BTRFS_WORK_HELPER(endio_helper);
97 BTRFS_WORK_HELPER(endio_meta_helper);
98 BTRFS_WORK_HELPER(endio_meta_write_helper);
99 BTRFS_WORK_HELPER(endio_raid56_helper);
100 BTRFS_WORK_HELPER(endio_repair_helper);
101 BTRFS_WORK_HELPER(rmw_helper);
102 BTRFS_WORK_HELPER(endio_write_helper);
103 BTRFS_WORK_HELPER(freespace_write_helper);
104 BTRFS_WORK_HELPER(delayed_meta_helper);
105 BTRFS_WORK_HELPER(readahead_helper);
106 BTRFS_WORK_HELPER(qgroup_rescan_helper);
107 BTRFS_WORK_HELPER(extent_refs_helper);
108 BTRFS_WORK_HELPER(scrub_helper);
109 BTRFS_WORK_HELPER(scrubwrc_helper);
110 BTRFS_WORK_HELPER(scrubnc_helper);
111 BTRFS_WORK_HELPER(scrubparity_helper);
112 
113 static struct __btrfs_workqueue *
114 __btrfs_alloc_workqueue(struct btrfs_fs_info *fs_info, const char *name,
115 			unsigned int flags, int limit_active, int thresh)
116 {
117 	struct __btrfs_workqueue *ret = kzalloc(sizeof(*ret), GFP_KERNEL);
118 
119 	if (!ret)
120 		return NULL;
121 
122 	ret->fs_info = fs_info;
123 	ret->limit_active = limit_active;
124 	atomic_set(&ret->pending, 0);
125 	if (thresh == 0)
126 		thresh = DFT_THRESHOLD;
127 	/* For low threshold, disabling threshold is a better choice */
128 	if (thresh < DFT_THRESHOLD) {
129 		ret->current_active = limit_active;
130 		ret->thresh = NO_THRESHOLD;
131 	} else {
132 		/*
133 		 * For threshold-able wq, let its concurrency grow on demand.
134 		 * Use minimal max_active at alloc time to reduce resource
135 		 * usage.
136 		 */
137 		ret->current_active = 1;
138 		ret->thresh = thresh;
139 	}
140 
141 	if (flags & WQ_HIGHPRI)
142 		ret->normal_wq = alloc_workqueue("btrfs-%s-high", flags,
143 						 ret->current_active, name);
144 	else
145 		ret->normal_wq = alloc_workqueue("btrfs-%s", flags,
146 						 ret->current_active, name);
147 	if (!ret->normal_wq) {
148 		kfree(ret);
149 		return NULL;
150 	}
151 
152 	INIT_LIST_HEAD(&ret->ordered_list);
153 	spin_lock_init(&ret->list_lock);
154 	spin_lock_init(&ret->thres_lock);
155 	trace_btrfs_workqueue_alloc(ret, name, flags & WQ_HIGHPRI);
156 	return ret;
157 }
158 
159 static inline void
160 __btrfs_destroy_workqueue(struct __btrfs_workqueue *wq);
161 
162 struct btrfs_workqueue *btrfs_alloc_workqueue(struct btrfs_fs_info *fs_info,
163 					      const char *name,
164 					      unsigned int flags,
165 					      int limit_active,
166 					      int thresh)
167 {
168 	struct btrfs_workqueue *ret = kzalloc(sizeof(*ret), GFP_KERNEL);
169 
170 	if (!ret)
171 		return NULL;
172 
173 	ret->normal = __btrfs_alloc_workqueue(fs_info, name,
174 					      flags & ~WQ_HIGHPRI,
175 					      limit_active, thresh);
176 	if (!ret->normal) {
177 		kfree(ret);
178 		return NULL;
179 	}
180 
181 	if (flags & WQ_HIGHPRI) {
182 		ret->high = __btrfs_alloc_workqueue(fs_info, name, flags,
183 						    limit_active, thresh);
184 		if (!ret->high) {
185 			__btrfs_destroy_workqueue(ret->normal);
186 			kfree(ret);
187 			return NULL;
188 		}
189 	}
190 	return ret;
191 }
192 
193 /*
194  * Hook for threshold which will be called in btrfs_queue_work.
195  * This hook WILL be called in IRQ handler context,
196  * so workqueue_set_max_active MUST NOT be called in this hook
197  */
198 static inline void thresh_queue_hook(struct __btrfs_workqueue *wq)
199 {
200 	if (wq->thresh == NO_THRESHOLD)
201 		return;
202 	atomic_inc(&wq->pending);
203 }
204 
205 /*
206  * Hook for threshold which will be called before executing the work,
207  * This hook is called in kthread content.
208  * So workqueue_set_max_active is called here.
209  */
210 static inline void thresh_exec_hook(struct __btrfs_workqueue *wq)
211 {
212 	int new_current_active;
213 	long pending;
214 	int need_change = 0;
215 
216 	if (wq->thresh == NO_THRESHOLD)
217 		return;
218 
219 	atomic_dec(&wq->pending);
220 	spin_lock(&wq->thres_lock);
221 	/*
222 	 * Use wq->count to limit the calling frequency of
223 	 * workqueue_set_max_active.
224 	 */
225 	wq->count++;
226 	wq->count %= (wq->thresh / 4);
227 	if (!wq->count)
228 		goto  out;
229 	new_current_active = wq->current_active;
230 
231 	/*
232 	 * pending may be changed later, but it's OK since we really
233 	 * don't need it so accurate to calculate new_max_active.
234 	 */
235 	pending = atomic_read(&wq->pending);
236 	if (pending > wq->thresh)
237 		new_current_active++;
238 	if (pending < wq->thresh / 2)
239 		new_current_active--;
240 	new_current_active = clamp_val(new_current_active, 1, wq->limit_active);
241 	if (new_current_active != wq->current_active)  {
242 		need_change = 1;
243 		wq->current_active = new_current_active;
244 	}
245 out:
246 	spin_unlock(&wq->thres_lock);
247 
248 	if (need_change) {
249 		workqueue_set_max_active(wq->normal_wq, wq->current_active);
250 	}
251 }
252 
253 static void run_ordered_work(struct __btrfs_workqueue *wq)
254 {
255 	struct list_head *list = &wq->ordered_list;
256 	struct btrfs_work *work;
257 	spinlock_t *lock = &wq->list_lock;
258 	unsigned long flags;
259 
260 	while (1) {
261 		void *wtag;
262 
263 		spin_lock_irqsave(lock, flags);
264 		if (list_empty(list))
265 			break;
266 		work = list_entry(list->next, struct btrfs_work,
267 				  ordered_list);
268 		if (!test_bit(WORK_DONE_BIT, &work->flags))
269 			break;
270 
271 		/*
272 		 * we are going to call the ordered done function, but
273 		 * we leave the work item on the list as a barrier so
274 		 * that later work items that are done don't have their
275 		 * functions called before this one returns
276 		 */
277 		if (test_and_set_bit(WORK_ORDER_DONE_BIT, &work->flags))
278 			break;
279 		trace_btrfs_ordered_sched(work);
280 		spin_unlock_irqrestore(lock, flags);
281 		work->ordered_func(work);
282 
283 		/* now take the lock again and drop our item from the list */
284 		spin_lock_irqsave(lock, flags);
285 		list_del(&work->ordered_list);
286 		spin_unlock_irqrestore(lock, flags);
287 
288 		/*
289 		 * We don't want to call the ordered free functions with the
290 		 * lock held though. Save the work as tag for the trace event,
291 		 * because the callback could free the structure.
292 		 */
293 		wtag = work;
294 		work->ordered_free(work);
295 		trace_btrfs_all_work_done(wq->fs_info, wtag);
296 	}
297 	spin_unlock_irqrestore(lock, flags);
298 }
299 
300 static void normal_work_helper(struct btrfs_work *work)
301 {
302 	struct __btrfs_workqueue *wq;
303 	void *wtag;
304 	int need_order = 0;
305 
306 	/*
307 	 * We should not touch things inside work in the following cases:
308 	 * 1) after work->func() if it has no ordered_free
309 	 *    Since the struct is freed in work->func().
310 	 * 2) after setting WORK_DONE_BIT
311 	 *    The work may be freed in other threads almost instantly.
312 	 * So we save the needed things here.
313 	 */
314 	if (work->ordered_func)
315 		need_order = 1;
316 	wq = work->wq;
317 	/* Safe for tracepoints in case work gets freed by the callback */
318 	wtag = work;
319 
320 	trace_btrfs_work_sched(work);
321 	thresh_exec_hook(wq);
322 	work->func(work);
323 	if (need_order) {
324 		set_bit(WORK_DONE_BIT, &work->flags);
325 		run_ordered_work(wq);
326 	}
327 	if (!need_order)
328 		trace_btrfs_all_work_done(wq->fs_info, wtag);
329 }
330 
331 void btrfs_init_work(struct btrfs_work *work, btrfs_work_func_t uniq_func,
332 		     btrfs_func_t func,
333 		     btrfs_func_t ordered_func,
334 		     btrfs_func_t ordered_free)
335 {
336 	work->func = func;
337 	work->ordered_func = ordered_func;
338 	work->ordered_free = ordered_free;
339 	INIT_WORK(&work->normal_work, uniq_func);
340 	INIT_LIST_HEAD(&work->ordered_list);
341 	work->flags = 0;
342 }
343 
344 static inline void __btrfs_queue_work(struct __btrfs_workqueue *wq,
345 				      struct btrfs_work *work)
346 {
347 	unsigned long flags;
348 
349 	work->wq = wq;
350 	thresh_queue_hook(wq);
351 	if (work->ordered_func) {
352 		spin_lock_irqsave(&wq->list_lock, flags);
353 		list_add_tail(&work->ordered_list, &wq->ordered_list);
354 		spin_unlock_irqrestore(&wq->list_lock, flags);
355 	}
356 	trace_btrfs_work_queued(work);
357 	queue_work(wq->normal_wq, &work->normal_work);
358 }
359 
360 void btrfs_queue_work(struct btrfs_workqueue *wq,
361 		      struct btrfs_work *work)
362 {
363 	struct __btrfs_workqueue *dest_wq;
364 
365 	if (test_bit(WORK_HIGH_PRIO_BIT, &work->flags) && wq->high)
366 		dest_wq = wq->high;
367 	else
368 		dest_wq = wq->normal;
369 	__btrfs_queue_work(dest_wq, work);
370 }
371 
372 static inline void
373 __btrfs_destroy_workqueue(struct __btrfs_workqueue *wq)
374 {
375 	destroy_workqueue(wq->normal_wq);
376 	trace_btrfs_workqueue_destroy(wq);
377 	kfree(wq);
378 }
379 
380 void btrfs_destroy_workqueue(struct btrfs_workqueue *wq)
381 {
382 	if (!wq)
383 		return;
384 	if (wq->high)
385 		__btrfs_destroy_workqueue(wq->high);
386 	__btrfs_destroy_workqueue(wq->normal);
387 	kfree(wq);
388 }
389 
390 void btrfs_workqueue_set_max(struct btrfs_workqueue *wq, int limit_active)
391 {
392 	if (!wq)
393 		return;
394 	wq->normal->limit_active = limit_active;
395 	if (wq->high)
396 		wq->high->limit_active = limit_active;
397 }
398 
399 void btrfs_set_work_high_priority(struct btrfs_work *work)
400 {
401 	set_bit(WORK_HIGH_PRIO_BIT, &work->flags);
402 }
403